Anda di halaman 1dari 9

Downloaded from http://jnnp.bmj.com/ on April 22, 2016 - Published by group.bmj.

com

JNNP Online First, published on March 23, 2016 as 10.1136/jnnp-2015-311890


Movement disorders

REVIEW

The prediagnostic phase of Parkinsons disease


Alastair John Noyce,1 Andrew John Lees,1 Anette-Eleonore Schrag2
Additional material is
published online only. To view
please visit the journal online
(http://dx.doi.org/10.1136/
jnnp-2015-311890).
1

Department of Molecular
Neuroscience, Reta Lila Weston
Institute for Neurological
Studies, UCL Institute of
Neurology, London, UK
2
Department of Clinical
Neuroscience, UCL Institute of
Neurology, London, UK
Correspondence to
Professor Anette Schrag,
Department of Clinical
Neuroscience, UCL Institute of
Neurology, Royal Free Campus,
University College London,
London NW3 2PF, UK; a.
schrag@ucl.ac.uk
Received 30 July 2015
Accepted 11 December 2015

ABSTRACT
The eld of prediagnostic Parkinsons disease (PD) is fast
moving with an expanding range of clinical and
laboratory biomarkers, and multiple strategies seeking to
discover those in the earliest stages or those at risk. It
is widely believed that the highest likelihood of securing
neuroprotective benet from drugs will be in these
subjects, preceding current point of diagnosis of PD. In
this review, we outline current knowledge of the
prediagnostic phase of PD, including an up-to-date
review of risk factors (genetic and environmental), their
relative inuence, and clinical features that occur prior to
diagnosis. We discuss imaging markers across a range of
modalities, and the emerging literature on uid and
peripheral tissue biomarkers. We then explore current
initiatives to identify individuals at risk or in the earliest
stages that might be candidates for future clinical trials,
what we are learning from these initiatives, and how
these studies will bring the eld closer to realistically
commencing primary or secondary preventive trials for
PD. Further progress in this eld hinges on greater
clinical and biological description, and understanding of
the prediagnostic, peridiagnostic and immediate
postdiagnostic stages of PD. Identifying subjects 3
5 years before they are currently diagnosed may be an
ideal group for neuroprotective trials. At the very least,
these initiatives will help clarify the stage before and
around diagnosis, enabling the eld to push into
unchartered territory at the earliest stages of disease.

INTRODUCTION

To cite: Noyce AJ, Lees AJ,


Schrag A-E. J Neurol
Neurosurg Psychiatry
Published Online First:
[please include Day Month
Year] doi:10.1136/jnnp2015-311890

The motor features of Parkinsons disease (PD)


(tremor, rigidity, slowness and balance problems)
are identied relatively late in the pathological
process when approximately 50% of dopaminergic
neurons have been lost in the substantia nigra.
Symptomatic treatment is efcacious, but there are
currently no drugs that demonstrably slow the
disease course. It is believed, albeit not proven, that
pathology may be too far advanced at the point of
clinical diagnosis to be affected by potentially neuroprotective treatments (assuming that these are
available). Identifying individuals at the earliest
stages of disease would pave the way for clinical
trials of emerging and repurposed drugs to prevent/
delay progression to clinically manifest PD (see
gure 1). However, modifying risk in those that do
not yet have a diagnosis represents a challenge. The
terms early disease or at-risk are frequently used
synonymously due to uncertainty about the point
at which the pathological process starts, but clarication will be important since it will determine
whether prevention is attempted on a primary or
secondary basis, and factors that initiate pathology
may not necessarily be the same as those that

subsequently drive progression. Inability to identify


disease activity before diagnosis precludes distinction of the two, but this limitation may be overcome, given current momentum in the eld of
biomarkers.
In this review, we describe current knowledge
and emerging ndings in the prediagnostic phase of
PD, including early features, genetic and environmental risk and protective factors, discuss current
strategies to identify individuals at earliest disease
stages to include in future clinical trials, and highlight how the knowledge gleaned from these
studies might bridge the gap into preventive or protective drug trials for PD.

IDENTIFICATION OF INDIVIDUALS AT THE


EARLIEST STAGES
Genetic factors
Having a family history of PD increases the odds of
PD by 34.5-fold, and up to 10% of patients
report a family history of PD.1 Studies into the
genetic basis of PD implicate lysosomal and mitochondrial dysfunction, and inammation in pathogenesis.2 3 Of the conrmed monogenic forms of
PD, most result in abnormalities of one or more of
these processes, but most are exceedingly rare and
do not account for elevated risk at a population
level (see gure 2). A central player in the disease is
-synuclein and mutations in the SNCA gene,
which encodes this protein, are a cause of familial
PD. Intraneuronal accumulation of -synuclein is
the pathological hallmark of PD, and mounting evidence suggests that brillar and oligomeric forms
of the protein may be neurotoxic.4 5 The full
picture of how these complex processes combine to
result in neurodegeneration remains incomplete,
but current theories include the possibility of prionlike cell-to-cell propagation.6
Mutations in the LRRK2 gene are the commonest
known genetic cause for PD, and the G2019S
mutation occurs in 4% of hereditary and 1% of
sporadic PD.7 LRRK2-related disease has agedependent penetrance (28% at 59 years, 51% at
69 years and 74% at 79 years), meaning that only a
proportion of carriers will develop PD during life.7
LRRK2 mutation carriers have been shown to have
subclinical dopaminergic abnormalities, measured
with functional imaging, and higher rates of nonmotor features of PD than non-carriers.8 9 In
patients with PD and LRRK2 mutations, the motor
picture may be similar to idiopathic PD, but wider
manifestations may not be.7 8
Heterozygous mutations in the glucocerebrosidase (GBA) gene are associated with an increased
risk of PD. Large studies have shown that GBA
mutations are common in Ashkenazi Jews,

Noyce AJ, et al. J Neurol Neurosurg Psychiatry 2016;0:18. doi:10.1136/jnnp-2015-311890

Copyright Article author (or their employer) 2016. Produced by BMJ Publishing Group Ltd under licence.

Downloaded from http://jnnp.bmj.com/ on April 22, 2016 - Published by group.bmj.com

Movement disorders

Figure 1 A schematic depicting normal (black solid line) and


Parkinsons disease-related (grey solid line) nigral cell loss over time,
including the point at diagnosis typically occurs (horizontal black
hashed line) and the potential for modifying the trajectory of
degeneration, if identied earlier (hashed grey line). NB. For colour
version grey lines appear red, black lines remain black.
occurring in 15% of patients and 3% of controls.10 In unselected patients with PD, 3.5% carry disease-associated GBA
mutations compared with <1% of controls.10 11 Clinical features of patients with PD and GBA mutations may be similar to
sporadic PD and generally respond to levodopa, but onset of
parkinsonism can be earlier, and carriers tend to have higher
rates of cognitive problems and R(apid)E(ye)M(ovement) sleep
behaviour disorder (RBD).11 Furthermore, early non-motor features of PD such as depression, subjective RBD and olfactory
dysfunction have been observed in carriers of GBA mutations
without PD, when compared with healthy controls.12
The study of manifesting and non-manifesting carriers of
LRRK2 and GBA mutations is important for understanding the
prodromal phase of PD, and for studies of drugs targeting specic pathways. Cohorts of these subjects have been assembled to
full these aims, but with greater study, further differences may
emerge between PD related to these mutations and sporadic

Figure 2 Risk factors and early features of Parkinsons disease


associated with increased (or decreased) risk of subsequent diagnosis.
Estimated magnitude of effect is plotted against estimated frequency.
SN+ is hyperechogenicity in the region of the substantia nigra using
transcranial sonography. Genetic and environmental risk factors are
shown in lighter grey, and prediagnostic features and hyperechogenicity
in darker grey. NB. For colour version genetic factors are red,
environmental factors are orange, pre-diagnostic features are green and
imaging features are blue.
2

disease. Other examples of monogenic PD include mutations,


duplications and triplications of the SNCA gene causing dominantly inherited PD, and mutations in PARK2, PINK1 and DJ1
causing recessively inherited forms of PD (for review see ref. 3),
with more emerging gradually. These monogenic causes of PD
are too rare to base predictive studies on, but they give important insight into disease mechanisms and therapeutic targets.
Mutations in single genes do not account for all the heritable
risk apparent in complex diseases, and genome-wide association
studies, in which large numbers of unrelated cases are compared
with unrelated controls, have yielded informative results. There
are at least 28 independent risk variants associated with PD that
increase or decrease risk in a small but potentially additive
way.13 This enables the construction polygenic risk proles,
pooling the combined effect of multiple variants to estimate risk
of PD or indeed age of disease onset.14 15 However, the heritable component of PD is estimated to be greater still, at around
30%, and identied risk loci and monogenic forms explain only
about 510%.16 Over time, with increasing numbers of studied
cases and controls, along with deep resequencing and precision
phenotyping, a greater proportion of the heritability of PD will
be uncovered. The inuence that genetic variation has on PD is
not limited to risk of getting disease alone, and specic variants
are likely to contribute additionally to age of disease onset, progression and phenotype, with a number of indicative studies
reported thus far.11 17 Furthermore, these and additional genetic
factors may dictate therapeutic choices in the future in the clinical setting and in terms of recruitment to clinical trials. The
genetic architecture of PD is continually expanding and increasingly complex. It has implications for multiple aspects of the
disease, but other factors are also important in determining risk.

Environmental risk factors


There is a large body of evidence demonstrating a small but signicantly elevated risk of PD associated with a number of environmental risk factors (see gure 2). Some of the strongest
environmental evidence exists for exposure to pesticide and
proxies for this, including farming occupation, rural living and
drinking well water.1 Specic implicated pesticides include rotenone and paraquat (structurally related to 1-methyl-4-phenyl1,2,3,6-tetrahydropyridine (MPTP), which has also been linked
to parkinsonism in users of illicit drugs18), and both these chemicals are used to create animal models of PD. Other potential
toxins include heavy metals such as manganese, with exposure
arising through occupations such as welding and in recreational
ephedrone users.19 It seems unlikely that environmental toxins
play more than a minor role in PD risk overall, with
meta-analysis suggesting ORs in the region of 1.22.0.1
Pooled observational study data also implicates head injury as
a minor but signicant risk factor for PD.1 20 There is increasing
evidence that individuals who suffer recurrent head injury, particularly sportspersons such as boxers, jockeys, American football and rugby players, are at risk of developing a range of
degenerative neurological conditions including parkinsonism,
dementia and motor neuron disease, although pathological
examination of these subjects tends to reveal alternative pathology than that typically associated with PD.21
By stark contrast with other common chronic diseases, there
exist a number of intriguing but consistent negative associations
with PD and lifestyle factors such as smoking, caffeine and
alcohol.1 Additionally, there are a number of medications for
which negative associations with PD have been reported in
observational studies, including calcium channel blockers,
non-steroidal anti-inammatories and statins.1 Some of these are

Noyce AJ, et al. J Neurol Neurosurg Psychiatry 2016;0:18. doi:10.1136/jnnp-2015-311890

Downloaded from http://jnnp.bmj.com/ on April 22, 2016 - Published by group.bmj.com

Movement disorders
currently the focus of clinical trials examining their potential
use in PD.
Whether these exposures offer true neuroprotective properties
or whether negative association, at least with lifestyle factors,
arises due to a common feature (eg, avoidance as part of an early
PD personality change) is yet to be determined. The former possibility is supported by clinical studies reporting improvement of
motor function in a clinical trial of caffeine to treat excessive
daytime somnolence in PD and PD animal models that show
protective effects of nicotine on nigrostriatal damage.22 23
Spurious association may arise as a result of reverse causality, a
potential problem with observational studies. This may be plausible even in prospective studies that exclude cases of incident PD
in the rst few years of follow-up because the prodromes of PD
are likely to be very long, during which time pathology is
present, but the classical features are not yet overt.
Another consistent negative association exists between levels
of serum urate and PD, with a number of studies demonstrating
a protective effect of elevated serum urate.1 The most recent
of these studies is an example of Mendelian randomisation
(MR), which is a powerful technique to assess casual relationships between risk factors and disease. In MR studies, a gene
variant is used as a proxy for an environmental exposure (or
intermediate phenotype), to determine the effect of this on a
disease outcome. Simon and colleagues used single nucleotide
polymorphisms in the SLC2A9 gene (which explain a proportion of the genetically specied variability in serum urate) to test
whether these polymorphisms predicted rate of progression in
early PD. They demonstrated that SLC2A9 status was associated
with elevated serum urate and was associated with slower
disease progression.24 The MR approach is protected against
biases that can arise in traditional observational studies assessing
causation; if, for example, reduction in serum urate occurred as
an early consequence of PD, a spurious negative association
between the exposure and outcome might arise. MR on the
other hand operates more like a randomised controlled trial in
which the exposure (in this case a gene variant) is randomly
allocated at baseline (conception) along with known and
unknown confounding factors. Given the consistency of this
relationship in the MR study with previous observational
studies, therapeutic modulation of urate levels is a strong target
for clinical trials.
Increasing research activity in exploring the overlap between
genes and the environment will further our understanding of
the causal basis of disease. As for many diseases, the total
picture of risk remains incomplete due to apparent and substantial randomness of the onset, the obscuration of risk factors
either because of rarity, ubiquity or poor measurement, or the
fact that disease tends to strike those at moderate risk, simply
because those at highest risk are far fewer.

Early clinical features


Recognition of the importance of non-motor features of PD has
been building for several years.25 26 In established PD, patients
regard non-motor symptoms at least as troublesome as motor
features, and treatment is often difcult.27 Non-motor symptoms are experienced early and there is substantial evidence
which suggests that they also predate diagnosis by several years
(see gure 2). A number of studies have demonstrated the association of PD with earlier diagnoses such as depression, anxiety,
constipation and erectile dysfunction.1 28 29
The best characterised early non-motor features are idiopathic
anosmia and RBD. Anosmia is relatively common in the ageing

population and is non-specic, with only a proportion going on


to develop neurodegenerative disease.30 RBD on the other
hand, is relatively specic for neurodegeneration, with high conversion rates, but clinical and pathological heterogeneity in that
it can precede PD, dementia with Lewy bodies and multiple
system atrophy.31 An important distinction lies between subjective RBD (on clinical history or questionnaire-based diagnosis)
and that which is formally diagnosed with an overnight sleep
study and polysomnography (PSG). PSG-conrmed RBD is rare
in the general population, and the largest observational study
amassed just over 300 subjects despite international collaborative efforts.32 Nonetheless study of PSG-conrmed RBD has
resulted in positive strides forward given that time during which
a proportion will convert to neurodegenerative disease is known
(25% at 3 years and 40% at 5 years).32 The emergence of
anosmia or subtle motor signs, in particular, in those with RBD
appears to further rene estimates of those that are likely to
convert.33 34
Characterisation of non-motor features is potentially valuable for early identication of PD and the Movement
Disorders Society (MDS) has recently released Research
Criteria for Prodromal PD.35 The literature initially described
the early phase of PD as being the pre-motor phase, but
more recently this has fallen out of favour with the recognition that subtle motor features can be present before diagnosis.28 36 The clinical diagnosis of PD requires multiple motor
features to be established, and while subtle motor signs may
be present, a clinical diagnosis of PD cannot be made until
they become more denite.4 Given that subtle or single motor
abnormalities occur prior to diagnosis and alongside early
non-motor features, this period is better referred to as the
prediagnostic phase.28 The MDS recommends the following
terminology:35
1. Preclinical PDpresence of neurodegenerative synucleinopathy without clinical symptoms (this stage will be dened by
disease biomarkers when available).
2. Prodromal PDpresence of early symptoms and signs
before PD diagnosis is possible.
3. Clinical PDdiagnosis of PD has been made based on the
presence of classical motor signs.
The emergence of large longitudinal primary care datasets has
allowed detailed exploration of the full range of early motor
and non-motor symptoms that predate PD, while being free
from the biases that many traditional observational studies have
suffered from.28 Alongside, advances in wearable technology,
and the availability of remote testing, will aid objective measurement of emerging motor dysfunction in those at risk of PD.37 A
variety of measuring devices have been developed, including
software applications which harness information on activity and
motion (and in some cases speech) captured by smart phones
and tablet-devices, custom-built sensors that measure gait, bradykinesia, dyskinesia and nocturnal movements, keyboard-based
speed and accuracy tests, as well as custom-built home-testing
devices (for examples of these, see online supplementary
material). Many of these devices demonstrate high sensitivity
and specicity in differentiating patients with PD from controls,
and are likely to see increasing use in clinical practice to guide
decision-making. Despite the indication that objective motor
dysfunction occurs prior to diagnosis in PD, currently there are
few examples of the application of remote or wearable devices
in prediagnostic PD. A major foreseeable hurdle is ensuring that
validation and regulatory approval for software and devices will
keep abreast of rapid and continuous change in available
technology.38

Noyce AJ, et al. J Neurol Neurosurg Psychiatry 2016;0:18. doi:10.1136/jnnp-2015-311890

Downloaded from http://jnnp.bmj.com/ on April 22, 2016 - Published by group.bmj.com

Movement disorders
Imaging markers
Radiotracer imaging with single photon emission computed
tomography (SPECT) or positron emission tomography (PET),
and transcranial sonography (TCS) have repeatedly demonstrated the ability to differentiate patients with PD from healthy
individuals with adequate sensitivity and specicity (see table
1).39 These modalities may also have the potential to identify
subclinical PD prior to diagnosis, and early support for this
notion came from studies showing that SPECT and PET were
typically abnormal bilaterally in patients with unilateral PD, and
could identify unaffected twins of patients with PD who later
went on to develop PD.39 Using SPECT, the clinical diagnosis of
PD tends to be made once there has been 4050% reduction in
tracer binding and an average 11.2% decline in striatal binding
has been observed each year after diagnosis.39 Multitracer PET
has also shown deterioration over time in patients with established PD, but is expensive when considering use on a large
scale.40
Evidence of demonstrable longitudinal change using DaT
(dopamine transporter) SPECT in early PD is now being sought
with serial studies performed as part of the Parkinsons
Progression Markers Initiative (PPMI) study.41 Furthermore,
progressive change in SPECT imaging has been reported in a
study of patients with RBD, and in the Parkinsons At Risk
Study (PARS), which includes subjects with idiopathic anosmia
(and other prodromal markers).42 43
By contrast, TCS demonstrating hyperechogenicity of the substantia nigra (SN) appears to be a static rather than changing
marker.39 A study in individuals with mild parkinsonian features, found a sensitivity of 91%, specicity of 82% and positive
predictive value (PPV) of 93% for SN hyperechogenicity and
PD diagnosis after follow-up, despite blinding those performing
sonography to clinical details at baseline.44 Current efforts to
improve standardisation and quantitative analysis for TCS and
SPECT may increase their utility in the prediagnostic phase of
disease, with SPECT more likely to demonstrate sensitivity to
change.
High-eld and novel sequences of MRI may also provide
opportunities to address the challenge of imaging disease progression in the prediagnostic phase of PD. Correlations of MRI
microstructural imaging abnormalities have been reported with
postmortem ndings and quantitative differences between
patients with PD and healthy subjects in terms of iron deposition, loss of neuromelanin and alterations in nigrosome 1, in
mainly single studies (for examples see online supplementary

material). In addition to subtle structural abnormalities, application of functional MRI methods have shown differences in functional connectivity with the basal ganglia network, the
default-mode network and the sensorimotor resting state
network, between patients with PD and healthy controls, some
of which are inuenced by dopaminergic medication (see online
supplementary material for examples of studies).
123
I-meta-iodobenzylguanidine (MIBG) myocardial scintigraphy has been largely studied in Japan, with multiple reports
showing a reduction in heart to mediastinum ratio of MIBG
uptake in patients with PD, compared with healthy controls or
other degenerative causes of parkinsonism.45 Cardiac sympathetic nerve involvement is a feature of incidental Lewy body
pathology, which is believed to be the pathological precursor to
PD.46 Altered MIBG uptake has been reported in patients with
a range of early non-motor features of PD, including autonomic
dysfunction, mood disorders and sleep disorders, meaning that
it may be a good prediagnostic imaging marker for PD, but
further studies are required.47

Fluid and tissue biomarkers


Given the supposed central importance of -synuclein to the
disease process, recent biomarker strategies have centred on
nding and characterising forms of the protein in a range of
biouids and tissues. Blood has been a disappointing target
to-date because red cells contain large quantities of -synuclein,
obscuring any theoretical difference in levels between patients
and controls, but some group differences have been demonstrated for DJ-1, urate, vitamin D and IGF-1 (for examples of
individual studies see online supplementary material).48 Plasma
apoliprotein A1 has been shown to differentiate patients from
healthy controls and, interestingly, also to correlate with DaT
SPECT decit in hyposmic subjects with a family history of PD,
recruited from the PARS study.49
It is believed that cerebrospinal uid (CSF) is most likely to
show changes representative of disease occurring in the brain.
Nonetheless, lumbar puncture is an invasive and costly procedure compared with blood draw, and is likely to remain most
appropriate for use in clinical trials and in specic clinical practice settings, unless a high-performing biomarker can be identied. Potential CSF biomarkers for PD include -synuclein and
DJ-1, with A42 potentially correlating with cognitive impairment, and various forms of tau protein and neurolament lightchain differentiating PD from atypical parkinsonian disorders.50
Separately, observed biomarker changes in saliva have included

Table 1 Imaging modalities and markers for Parkinsons disease


Modality

Example tracers/sequence

Observation(s)

Analysis

Accessibility

Cost

Suitability
for screening

SPECT

123

Loss of binding in striatum

Visual inspection, quantification

++

+++

+++

Hyperechogenicity in the region of


the substantia nigra
Loss of binding in striatum
May help differentiate atypical PD
Numerous reported, none
established

Visual inspection, quantification

++++

+++

Visual inspection, quantification

++

++++

Visual inspection, quantification

+++

++

++

Low heart-to-mediastinum ratio

Visual inspection, quantification

++

++

++

I--CIT
I-FP-CIT
23.5 Hz transducer

123

TCS
PET

18

F-dopa
F-FDG
Traditional (T1& T2), T2/T2*
(gradient echo), DTI, spin
echo, fMRI
123
I-meta-iodobenzylguanidine
18

MRI

MIBG

Accessibility, cost and suitability for screening are estimated semiquantitatively on a four-point relative scale with + lowest and ++++ highest.
DTI, diffusion tensor imaging; FDG, fludeoxyglucose; fMRI, functional MRI; Hz, hertz; MIBG, metaiodobenzylguanidine; PET, positron emission tomography; SPECT, single photon
emission computed tomography; TCS, transcranial sonography.

Noyce AJ, et al. J Neurol Neurosurg Psychiatry 2016;0:18. doi:10.1136/jnnp-2015-311890

Downloaded from http://jnnp.bmj.com/ on April 22, 2016 - Published by group.bmj.com

Movement disorders
reduction in -synuclein levels and elevation of DJ-1 in the
saliva (see online supplementary material for examples of individual studies of CSF and salivary biomarkers).
-Synuclein pathology is found outside the central nervous
system in patients with PD (even preceding diagnosis in some
instances), using a variety of sampling methods, immunohistochemical techniques, in fresh and archival tissue. The gut has
been proposed as the site of initiation of PD pathology, but the
burden of synucleinopathy correlates poorly with disease severity and the proximal (more) to distal (less) gradient in pathology
appears at odds with this suggestion.51 52 In addition, there is
variability in the reporting of cell loss and a current lack of consensus of what should be classied as Lewy pathology given that
the appearances in gut are dissimilar to what is seen typically in
the brain of patients, and the fact that control subjects have variable staining for -synuclein.53 The potential for gut biomarkers
remains high, however, and may not be restricted to tissue sampling. Recently, interest has been stirred by demonstrable differences in the faecal microbiome of patients and controls.54
Salivary glands may also be a peripheral source for demonstrating PD-related pathology but, like gut biopsies, the invasive
nature and cost of deep tissue sampling or full colonoscopy currently make both undesirable.52 55 Of signicant interest is a
recent study showing synuclein deposition in dermal punch
biopsies taken from patients with PD, and differentiating these
subjects from other forms of parkinsonism.56 While these ndings require replication, the skin offers the exciting possibility of
yielding biomarkers for PD, and is signicantly more accessible
than salivary glands and the gut. A summary of tissue and uid
biomarkers is provided in table 2.

CURRENT STUDIES MAPPING THE PREDIAGNOSTIC AND


PERIDIAGNOSTIC PHASES OF PD
Several studies (table 3) have been initiated: to identify those in
the prediagnostic and prodromal phases of PD; identify clinical
and biological markers to track progression of disease before
diagnosis; create platforms to nd subjects for inclusion in neuroprotective drug trials. As outlined above, some initiatives
recruit individuals with a single strong risk factor such as carrier
status for LRRK2 or GBA mutations, or idiopathic RBD or
anosmia, in order that subjects may be followed prospectively,
whereas other approaches employ large population-based
cohorts or retrospective casecontrol methods to examine associations with PD and preceding diagnoses. From the former we
learn more about the emergence of PD in specic risk cohorts,
which, in turn, may prove to be appropriate for recruitment to
clinical trials. They are likely to be more homogenous in terms
of their disease mechanisms, pathology and clinical features, as
well as being the simplest in which to determine time to

conversion. However, they are perhaps not representative of


the spectrum of PD as a whole, and biomarkers developed in
these groups must be replicated in sporadic PD cohorts before
assuming that they are applicable to all. The latter are difcult,
and costly to conduct with in-depth assessments and appropriate
sample sizes, but allow the investigation of risk/protective
factors and early symptoms and signs that precede emergence of
established PD. This, in turn, informs strategic combination of
factors to try and delineate individuals at high risk while also
capturing the full spectrum of PD. Although the magnitude of
risk associated with individual risk factors and early non-motor
features has been reported, the best combination of risk factors
for predicting PD remains unknown.1 35 Several studies are now
seeking to combine risk factors for PD in order to improve predictive power with which those at increased risk of PD can be
identied, with and without imaging markers.
The Prospective validation of Risk factors for the development of Parkinson Syndromes (PRIPS) study was a large prospective study that sought to determine the magnitude of risk of
PD that SN hyperechogenicity conveyed.57 Over 1800 participants were screened and 304 had hyperechogencity. At 3 years
follow-up, 11 had developed PD and the relative risk of incident
PD was 17.3 (95% CI 3.7 to 81.3) if there was SN hyperechogenicity at baseline.57 The aforementioned PARS study uses
objective smell testing to identify subjects with idiopathic
anosmia at stage 1, followed by DaT SPECT at stage 2 to identify subclinical presynaptic denervation.43 The study has
reported early results which demonstrated reductions in nigrostriatal DaT binding in subjects with hyposmia compared to
those with normal smell, as well as associations between a
number of prodromal features of PD within the study cohort.43
The Tbinger evaluation of Risk factors for the Early detection
of NeuroDegeneration (TREND) study examines subjects who
are over 50 years of age, with a limited combination of risk
factors (anosmia or depression or RBD), using serial studies of
movement, laboratory tests and imaging, with follow-up to incident PD. Baseline data from this cohort have been reported
showing associations between selected prodromal markers and
other early associated features of PD.58
Two large multicentre studies, one coordinated from the USA
(the Parkinsons Progression Markers Initiative, PPMI) and one
based in the UK (the Tracking Parkinsons or PRoBaND study),
recruit patients immediately after the clinical diagnosis of PD
and undertake detailed clinical, imaging and biomarker studies
longitudinally. While not strictly looking at prediagnostic PD,
the PPMI and PRoBaND studies will help dene the role of
clinical markers (motor and non-motor) in the early measurement of PD, and the identication of novel imaging and laboratory biomarkers, as well as giving insight into what might be

Table 2 Tissue and fluid biomarkers for PD


Source

Medium

Main techniques

Candidate markers

Invasiveness

Cost

Suitability
for screening

Cerebrospinal fluid
Blood
Saliva
Gut
Salivary gland
Skin

Fluid
Fluid
Fluid
Tissue biopsy
Tissue biopsy
Tissue biopsy

Assay, electrophoresis, mass-spec


Assay, electrophoresis, mass-spec
Assay, electrophoresis, mass-spec
Immunohistochemistry
Immunohistochemistry
Immunohistochemistry

t--syn, p--syn, t-tau, p-tau, NFL


Uric acid, inflammatory markers, IGF-1, ApoA1, DJ-1
-syn, DJ-1
t--syn, p--syn
t--syn, p--syn
t--syn, p--syn

+++
++
+
++++
++++
+++

++
+
+
+++
+++
++

++
+++
+++
++
+
+++

Invasiveness, cost and suitability for screening are estimated semiquantitatively on a four-point relative scale with + lowest and ++++ highest.
ApoA1, apolipoprotein A1; IGF-1, insulin like growth factor 1; NFL, neurofilament light chain; p--syn, phosphorylated synuclein; t--syn, total synuclein.

Noyce AJ, et al. J Neurol Neurosurg Psychiatry 2016;0:18. doi:10.1136/jnnp-2015-311890

Downloaded from http://jnnp.bmj.com/ on April 22, 2016 - Published by group.bmj.com

Movement disorders
Table 3 Studies in the prediagnostic phase of Parkinsons disease
Acronym

Number
recruited

Tests/exposures

Outcome

Germany,
Austria

1847

TCS, smell, UPDRS

Clinical diagnosis
of PD

Subjects over 50 years with


anosmia, self-report RBD or
depression

Germany

>1200

Clinical diagnosis
of PD

Subjects over 50 years with


hyposmia and DaT deficit on
SPECT
Subjects with prodromal
features or gene mutations

USA

4999 (completed
baseline smell
test)
Anosmia/RBD=65
Genetic=150

TCS, smell, UPDRS,


quantitative motor,
psychometry, blood
biomarkers
Smell, DaT SPECT, UPDRS,
cognition, blood biomarkers

Subjects with a first-degree


relative with PD, or subjects
with RBD (PSG confirmed)
Subjects over 60 years old

UK

190

UK

1323

Study name

Participants

Country

Prospective validation of risk


factors for the development of
Parkinson syndromes
Tbinger evaluation of risk
factors for the early detection of
neurodegeneration

Subjects over 50 years old

PARS43

Parkinsons at-risk syndrom


study

P-PPMI41

Prodromal Parkinsons
Progression Markers Initiative

OPDC59

Oxford Parkinsons Disease


Centre study

PREDICT-PD60

PREDICT-PD

PRIPS

57

TREND58

International

CSF and blood biomarkers,


UPDRS, cognition, sleep and
autonomic assessments
UPDRS, non-motor
assessments, blood and CSF
biomarkers
In all (online): risk scoring,
smell, RBDSQ, quantitative
motor (BRAIN test), genetics
In extremes of risk: UPDRS,
cognitive, TCS

Clinical diagnosis
of PD/DaT deficit
on SPECT
Clinical diagnosis
of PD
Clinical diagnosis
of PD
Clinical diagnosis
of PD/DaT deficit
on SPECT

BRAIN test, BRadykinesia Akinesia INcoordination test; B-SIT, Brief Smell Identification Test; CSF, Cerebrospinal Fluid; DaT, Dopamine Transporter; PD, Parkinsons disease; PSG,
polysomnography; RBD, REM sleep Behaviour Disorder; RBDSQ, RBD Screening Questionnaire; SPECT, single photon emission computed tomography; TCS, Transcranial Sonography;
UPDRS, Unified Parkinsons Disease Rating Scale.

apparent through back-extrapolation to the prediagnostic phase.


The PPMI study also includes a prodromal arm (P-PPMI) in
which subjects with RBD, anosmia or a mutation (LRRK2, GBA
or SNCA), will be assessed and followed in the same way as PD
subjects, allowing for a seamless examination of the prediagnostic and early disease stages of PD. Separately, as part of a large
study aimed at understanding the biological basis of disease in
patients with established PD, the Oxford Parkinsons Disease
Centre (OPDC) includes smaller high-risk groups with a family
history or RBD. Clinical assessments, laboratory and imaging
biomarker studies are being undertaken and early results are
emerging.59
In the UK, the PREDICT-PD study combines risk factors and
early non-motor features to devise a risk scoring process for
future PD. Risk scores were calculated based on a meta-analysis
of the published literature.1 These, in turn, were used to generate ORs for the effect on risk of PD ascribed by individual early

non-motor features and risk factors. Using a priori odds of PD


attributed to age, a Bayesian model of risk was constructed to
yield combined risk estimates for each subject in the study.60
The study runs almost entirely via the internet with more
detailed laboratory, motor and imaging investigation for groups
at the extremes of risk. PREDICT-PD is the rst study to try
and combine large numbers of risk factors for PD and has the
potential to screen a large, community-based population, and
aims to facilitate recruitment into clinical trials in the future.
Unlike some of the other studies, it seeks to identify individuals
spanning the full spectrum of PD, which makes this cohort
highly applicable to occurrence of typical PD in clinical settings.

FURTHER CHALLENGES AND OPPORTUNITIES IN THE


PREDIAGNOSTIC PHASE
The above studies aim to overcome the important challenge of
identication of at-risk individuals who may develop the

Figure 3 A schematic showing


determinants of risk, the prediagnostic
phase ( preclinical and prodromal
phases) and clinical phase of
Parkinsons disease, along with the
parallel application of risk and disease
progression markers to measure
disease activity across phases.

Noyce AJ, et al. J Neurol Neurosurg Psychiatry 2016;0:18. doi:10.1136/jnnp-2015-311890

Downloaded from http://jnnp.bmj.com/ on April 22, 2016 - Published by group.bmj.com

Movement disorders
classical clinical syndrome of PD, with the eventual aim of initiating treatment to avoid or delay clinically relevant symptoms.
In addition, studies such as PPMI (with P-PPMI), TREND,
PARS and PREDICT-PD will document the time immediately
before, during and after the emergence of clinically recognisable
PD, delineating the clinical and biomarker features of this phase
that will be crucial to commencing clinical trials. These studies
will help rene the determination of risk status and course of
early disease progression (see gure 3).
There are additional hurdles that must be overcome before
clinical trials recruiting subjects in this phase of disease can be
planned: (1) determination of appropriate study endpoints and
duration of trials. Prevention or delaying emergence of classical
symptom onset is the ultimate aim, but PD is insidious, with its
clinical manifestations emerging over months and years, making
many clinical end points unsuitable and such studies difcult to
fund for the duration required. A sensitive clinical marker of
progression would be valuable in detecting subtle changes at
this early phase, however, an imaging or laboratory marker that
spanned the prediagnostic and early postdiagnostic phases may
offer better sensitivity, specicity, reliability and precision overall
(see gure 3). This, in turn, could allow appropriate calculation
of sample sizes and trial duration, dependent on anticipated
drug effect; (2) another important consideration in trial design
is the heterogeneity in clinical manifestation of PD, rate of progression and the presence/absence of other features. Clinical
trials designed to show the disease-modifying effect of an agent
may initially need to include homogenous samples or samples
stratied for presentation and rate of progression in order to
show an effect before trials in wider groups can be conducted;
(3) other factors are continuity and applicability through the
early stages of the disease. Even with an optimised early detection process, there will still be individuals who are undetected,
and rst present with overt signs of PD, and potential treatments will need to be assessed for demonstrable effects in these
subjects too. Longer-term observational studies that examine
risk status could support registries through which subjects indicate their willingness to participate in future clinical trials and
biomarker initiatives. Consenting eligible subjects could be
offered inclusion into clinical trials with the benet of extensive
available pre-trial data, but issues of selection bias and generalisability of results must be considered; (4) Finally, there are
ethical implications of treating at-risk populations. For a repurposed drug, with previous data on safety and tolerance, the
implications of undertaking clinical trials in those at-risk are
perhaps less than for novel drugs with unknown safety proles
and potential toxicity. Justication for more invasive therapies
could probably not be found without clear results in established
PD. In addition, disclosure of risk status is likely to be a prerequisite for participation in clinical trials, but has the potential
to bias recruitment and poses an ethical barrier in the absence
of proven neuroprotective effects. Ultimately, disclosure may be
unavoidable in order to make an informed decision about trial
participation.

major interest is whether these differences can be demonstrated


in high-risk/prediagnostic subjects, and whether they change up
to the point of diagnosis and immediately beyond. This will
enable testing of drug therapies at a time when more neuronal
tissue can potentially be preserved, and there is an absence of
symptomatic effects of medication with the potential to
confound.

CONCLUSION

15

Signicant progress has been made in the understanding and


identication of subjects in the prediagnostic phase of PD and a
number of initiatives are underway to further dene these
groups. These studies may contain subjects that would be candidates for recruitment into clinical trials targeting neuroprotection within a few years. Parallel exploration of peripheral tissue,
uid and multimodal imaging is needed to identify differences
between patients and controls across a range of markers. Of

Contributors AJN conceived the topic and drafted the manuscript. AJL provided
critical revision. AS conceived the topic and provided critical revision.
Funding This work was supported by Parkinsons UK (career development award
for AJN, reference F-1201).
Competing interests AJN has received grant money from lan/Prothena
Pharmaceuticals and from GE Healthcare, and honoraria from Ofce Octopus. AJL
has received honoraria from Novartis, Teva, Meda, Boehringer Ingelheim, GSK,
Ipsen, Lundbeck, Allergan and Orion. AS has received grant money from GE
Healthcare and honoraria from UCB.
Provenance and peer review Not commissioned; externally peer reviewed.
Data sharing statement No additional data are available.
Open Access This is an Open Access article distributed in accordance with the
terms of the Creative Commons Attribution (CC BY 4.0) license, which permits
others to distribute, remix, adapt and build upon this work, for commercial use,
provided the original work is properly cited. See: http://creativecommons.org/
licenses/by/4.0/

REFERENCES
1
2
3
4
5
6

8
9
10
11

12

13

14

16

17

18

Noyce AJ, et al. J Neurol Neurosurg Psychiatry 2016;0:18. doi:10.1136/jnnp-2015-311890

Noyce AJ, Bestwick JP, Silveira-Moriyama L, et al. Meta-analysis of early nonmotor


features and risk factors for Parkinson disease. Ann Neurol 2012;72:893901.
Schapira AH, Jenner P. Etiology and pathogenesis of Parkinsons disease.
Mov Disord 2011;26:104955.
Singleton AB, Farrer MJ, Bonifati V. The genetics of Parkinsons disease: progress
and therapeutic implications. Mov Disord 2013;28:1423.
Gibb WR, Lees AJ. The relevance of the Lewy body to the pathogenesis of
idiopathic Parkinsons disease. J Neurol Neurosurg Psychiatry 1988;51:74552.
Peelaerts W, Bousset L, Van der Perren A, et al. -Synuclein strains cause distinct
synucleinopathies after local and systemic administration. Nature 2015;522:3404.
McCann H, Cartwright H, Halliday GM. Neuropathology of -synuclein propagation
and Braak hypothesis. Mov Disord. Published Online First: 4 Sep 2015.
doi10.1002/mds.26421
Healy DG, Falchi M, OSullivan SS, et al. Phenotype, genotype, and worldwide
genetic penetrance of LRRK2-associated Parkinsons disease: a case-control study.
Lancet Neurol 2008;7:58390.
Marras C, Schle B, Schuele B, et al. Phenotype in parkinsonian and
nonparkinsonian LRRK2 G2019S mutation carriers. Neurology 2011;77:32533.
Nandhagopal R, McKeown MJ, Stoessl AJ. Functional imaging in Parkinson disease.
Neurology 2008;70(16 Pt 2):147888.
Sidransky E, Nalls MA, Aasly JO, et al. Multicenter analysis of glucocerebrosidase
mutations in Parkinsons disease. N Engl J Med 2009;361:165161.
Winder-Rhodes SE, Evans JR, Ban M, et al. Glucocerebrosidase mutations inuence
the natural history of Parkinsons disease in a community-based incident cohort.
Brain 2013;136:3929.
Beavan M, McNeill A, Proukakis C, et al. Evolution of prodromal clinical markers of
Parkinson disease in a GBA mutationpositive cohort. JAMA Neurol
2015;72:2018.
Nalls MA, Pankratz N, Lill CM, et al. Large-scale meta-analysis of genome-wide
association data identies six new risk loci for Parkinsons disease. Nat Genet
2014;46:98993.
Escott-Price V, Nalls MA, Morris HR, et al., International Parkinsons Disease
Genomics Consortium. Polygenic risk of Parkinson disease is correlated with disease
age at onset. Ann Neurol 2015;77:58291.
Nalls MA, McLean CY, Rick J, et al. Diagnosis of Parkinsons disease on the basis
of clinical and genetic classication: a population-based modelling study. Lancet
Neurol 2015;14:10029.
Keller MF, Saad M, Bras J, et al. Using genome-wide complex trait analysis to
quantify missing heritability in Parkinsons disease. Hum Mol Genet
2012;21:49965009.
Ritz B, Rhodes SL, Bordelon Y, et al. -Synuclein genetic variants predict faster
motor symptom progression in idiopathic Parkinson disease. PLoS ONE 2012;7:
e36199.
Langston JW, Ballard P, Tetrud JW, et al. Chronic Parkinsonism in humans due to a
product of meperidine-analog synthesis. Science 1983;219:97980.

Downloaded from http://jnnp.bmj.com/ on April 22, 2016 - Published by group.bmj.com

Movement disorders
19

20
21
22
23
24
25
26

27
28
29

30
31

32

33
34

35
36

37
38

39
40

Selikhova M, Fedoryshyn L, Matviyenko Y, et al. Parkinsonism and dystonia caused


by the illicit use of ephedronea longitudinal study. Mov Disord
2008;23:222431.
Gardner RC, Burke JF, Nettiksimmons J, et al. Traumatic brain injury in later life
increases risk for Parkinson disease. Ann Neurol 2015;77:98795.
McKee AC, Stein TD, Nowinski CJ, et al. The spectrum of disease in chronic
traumatic encephalopathy. Brain 2013;136:4364.
Postuma RB, Lang AE, Munhoz RP, et al. Caffeine for treatment of Parkinson
disease: a randomized controlled trial. Neurology 2012;79:6518.
Quik M, Perez XA, Bordia T. Nicotine as a potential neuroprotective agent for
Parkinsons disease. Mov Disord 2012;27:94757.
Simon KC, Eberly S, Gao X, et al. Mendelian randomization of serum urate and
Parkinson disease progression. Ann Neurol 2014;76:8628.
Chaudhuri KR, Healy DG, Schapira AHV, et al. Non-motor symptoms of Parkinsons
disease: diagnosis and management. Lancet Neurol 2006;5:23545.
Gallagher DA, Lees AJ, Schrag A. What are the most important nonmotor
symptoms in patients with Parkinsons disease and are we missing them? Mov
Disord 2010;25:2493500.
Chaudhuri KR, Schapira AH. Non-motor symptoms of Parkinsons disease:
dopaminergic pathophysiology and treatment. Lancet Neurol 2009;8:46474.
Schrag A, Horsfall L, Walters K, et al. Prediagnostic presentations of Parkinsons
disease in primary care: a case-control study. Lancet Neurol 2015;14:5764.
Adams-Carr KL, Bestwick JP, Shribman S, et al. Constipation preceding
Parkinsons disease: a systematic review and meta-analysis. J Neurol Neurosurg
Psychiatry 2015.
Hawkes C. Olfaction in neurodegenerative disorder. Mov Disord 2003;18:36472.
Iranzo A, Molinuevo JL, Santamara J, et al. Rapid-eye-movement sleep behaviour
disorder as an early marker for a neurodegenerative disorder: a descriptive study.
Lancet Neurol 2006;5:5727.
Postuma RB, Iranzo A, Hogl B, et al. Risk factors for neurodegeneration in
idiopathic rapid eye movement sleep behavior disorder: a multicenter study. Ann
Neurol 2015;77:8309.
Mahlknecht P, Iranzo A, Hogl B, et al. Olfactory dysfunction predicts early transition
to a Lewy body disease in idiopathic RBD. Neurology 2015;84:6548.
Postuma RB, Gagnon J-F, Bertrand J-A, et al. Parkinson risk in idiopathic REM sleep
behavior disorder: preparing for neuroprotective trials. Neurology
2015;84:110413.
Berg D, Postuma RB, Adler CH, et al. MDS research criteria for prodromal
Parkinsons disease. Mov Disord 2015;30:160011.
Postuma RB, Lang AE, Gagnon JF, et al. How does parkinsonism start? Prodromal
parkinsonism motor changes in idiopathic REM sleep behaviour disorder. Brain
2012;135:186070.
Maetzler W, Domingos J, Srulijes K, et al. Quantitative wearable sensors for
objective assessment of Parkinsons disease. Mov Disord 2013;28:162837.
Pasluosta CF, Gassner H, Winkler J, et al. An emerging era in the management of
Parkinsons disease: wearable technologies and the internet of things. IEEE J
Biomed Health Inform 2015;19:187381.
Brooks DJ. Imaging approaches to Parkinson disease. J Nucl Med
2010;51:596609.
Nandhagopal R, Kuramoto L, Schulzer M, et al. Longitudinal progression of
sporadic Parkinsons disease: a multi-tracer positron emission tomography study.
Brain 2009;132:29709.

41
42

43
44

45

46

47
48
49
50

51

52

53

54
55
56
57

58
59

60

Marek K, Jennings D, Lasch S, et al. The Parkinson Progression Marker Initiative


(PPMI). Prog Neurobiol 2011;95:62935.
Iranzo A, Valldeoriola F, Lomea F, et al. Serial dopamine transporter imaging of
nigrostriatalfunction in patients with idiopathic rapid-eye-movement sleep behaviour
disorder: a prospective study. Lancet Neurol 2011;10:797805.
Jennings D, Siderowf A, Stern M, et al. Imaging prodromal Parkinson disease: the
Parkinson Associated Risk Syndrome study. Neurology 2014;83:173946.
Gaenslen A, Unmuth B, Godau J, et al. The specicity and sensitivity of transcranial
ultrasound in the differential diagnosis of Parkinsons disease: a prospective blinded
study. Lancet Neurol 2008;7:41724.
Orimo S, Suzuki M, Inaba A, et al. 123I-MIBG myocardial scintigraphy for
differentiating Parkinsons disease from other neurodegenerative parkinsonism: a
systematic review and meta-analysis. Parkinsonism Relat Disord 2012;18:
494500.
Orimo S, Takahashi A, Uchihara T, et al. Degeneration of cardiac sympathetic nerve
begins in the early disease process of Parkinsons disease. Brain Pathol
2007;17:2430.
Sakakibara R, Tateno F, Kishi M, et al. MIBG myocardial scintigraphy in pre-motor
Parkinsons disease: a review. Parkinsonism Relat Disord 2014;20:26773.
Chahine LM, Stern MB, Chen-Plotkin A. Blood-based biomarkers for Parkinsons
disease. Parkinsonism Relat Disord 2014;20:S99103.
Qiang JK, Wong YC, Siderowf A, et al. Plasma apolipoprotein A1 as a biomarker
for Parkinson disease. Ann Neurol 2013;74:11927.
Magdalinou N, Lees AJ, Zetterberg H. Cerebrospinal uid biomarkers in
parkinsonian conditions: an update and future directions. J Neurol Neurosurg
Psychiatry 2014;85:106575.
Lebouvier T, Neunlist M, Bruley des Varannes S, et al. Colonic biopsies to assess the
neuropathology of Parkinsons disease and its relationship with symptoms. PLoS
ONE 2010;5:e12728.
Beach TG, Adler CH, Sue LI, et al., Arizona Parkinsons Disease Consortium.
Multi-organ distribution of phosphorylated -synuclein histopathology in subjects
with Lewy body disorders. Acta Neuropathol 2010;119:689702.
Visanji NP, Marras C, Hazrati LN, et al. Alimentary, my dear Watson? The
challenges of enteric -synuclein as a Parkinsons disease biomarker. Mov Disord
2014;29:44450.
Keshavarzian A, Green SJ, Engen PA, et al. Colonic bacterial composition in
Parkinsons disease. Mov Disord 2015;30:135160.
Beach TG, Adler CH, Dugger BN, et al. Submandibular gland biopsy for the
diagnosis of Parkinson disease. J Neuropathol Exp Neurol 2013;72:1306.
Donadio V, Incensi A, Leta V, et al. Skin nerve -synuclein deposits: a biomarker for
idiopathic Parkinson disease. Neurology 2014;82:13629.
Berg D, Seppi K, Behnke S, et al. Enlarged substantia nigra hyperechogenicity and
risk for Parkinson disease: a 37-month 3-center study of 1847 older persons. Arch
Neurol 2011;68:9327.
Gaenslen A, Wurster I, Brockmann K, et al. Prodromal features for Parkinsons
diseasebaseline data from the TREND study. Eur J Neurol 2014;21:76672.
Baig F, Lawton M, Rolinski M, et al. Delineating nonmotor symptoms in early
Parkinsons disease and rst-degree relatives. Mov Disord. Published Online First:
14 Jul 2015. doi:10.1002/mds.26281
Noyce AJ, Bestwick JP, Silveira-Moriyama L, et al. PREDICT-PD: Identifying risk of
Parkinsons disease in the community: methods and baseline results. J Neurol
Neurosurg Psychiatry 2013;85:317.

Noyce AJ, et al. J Neurol Neurosurg Psychiatry 2016;0:18. doi:10.1136/jnnp-2015-311890

Downloaded from http://jnnp.bmj.com/ on April 22, 2016 - Published by group.bmj.com

The prediagnostic phase of Parkinson's


disease
Alastair John Noyce, Andrew John Lees and Anette-Eleonore Schrag
J Neurol Neurosurg Psychiatry published online January 11, 2016

Updated information and services can be found at:


http://jnnp.bmj.com/content/early/2016/03/23/jnnp-2015-311890

These include:

Supplementary Supplementary material can be found at:


Material http://jnnp.bmj.com/content/suppl/2016/01/11/jnnp-2015-311890.DC1.
html

References

This article cites 57 articles, 9 of which you can access for free at:
http://jnnp.bmj.com/content/early/2016/03/23/jnnp-2015-311890#BIBL

Open Access

This is an Open Access article distributed in accordance with the terms of


the Creative Commons Attribution (CC BY 4.0) license, which permits
others to distribute, remix, adapt and build upon this work, for commercial
use, provided the original work is properly cited. See:
http://creativecommons.org/licenses/by/4.0/

Email alerting
service

Receive free email alerts when new articles cite this article. Sign up in the
box at the top right corner of the online article.

Topic
Collections

Articles on similar topics can be found in the following collections


Open access (203)
Drugs: CNS (not psychiatric) (1856)
Parkinson's disease (663)
Stroke (1403)

Notes

To request permissions go to:


http://group.bmj.com/group/rights-licensing/permissions
To order reprints go to:
http://journals.bmj.com/cgi/reprintform
To subscribe to BMJ go to:
http://group.bmj.com/subscribe/